The Great Storm that struck south-east England and northern France in October 1987 left at least 22 people dead. What made the storm so exceptionally damaging – and, as a meteorologist, so exceptionally interesting – is that it was accompanied by a phenomenon we now know as a sting jet.
Mid-latitude cyclones like the Great Storm and most other major storms that hit the UK tend to extend over several hundred kilometres, and they can generate strong surface winds over a wide area. The sting jet, on the other hand, is a narrow current of air within some such cyclones which accelerates as it descends into the lower atmosphere. There it produces potentially the most severe wind damage over a swath usually less than 100 kilometres across.
Fifteen years ago the existence of sting jets was unknown. But since then various scientists, including myself, have studied them and we are beginning to understand them better. This work is important as it is still quite a challenge to factor such small-scale features into weather forecasts, however lethal their effects may be.
In search of the sting jet
One could be forgiven for believing that weather science was sufficiently mature that most significant phenomena would already have been identified. So it was exciting to find this new feature, which wasn’t just a more severe version of the things we knew about but, rather, was different in kind and not just in degree.
The first step of course was to identify that such a phenomenon exists: this came about as a result of an especially detailed analysis of the October 1987 storm which was published in 2004.
Important clues pointing to the existence of the new phenomenon were provided by satellite cloud imagery. The strongest winds were seen to emerge at the tip of a hook of cloud curving around the cyclone centre, much like a scorpion’s sting is found at the end of its curved tail. We therefore named it a “sting jet”.
The wind speed in the sting jet far exceeded the rate of advance of the cloud tip. This indicated that the dry air in the red area could only have come from within this cloud. The cloud was in fact evaporating because of the warming caused by the descent of the sting jet. The amount of warming was, however, mitigated by virtue of some cooling due to evaporation.
This cooling was thought to play a role in bringing the sting jet winds down to the surface. Another characteristic of the cloud in this region was that it was found to be composed of a series of cloud fingers. This kind of structure is consistent with a particular form of rather small-scale instability. It is the resulting small-scale sloping layers of ascending and descending air that give rise to the observed fingers of cloud.
These findings triggered a number of studies that confirmed that sting jets do descend as they accelerate and that the suspected small-scale dynamical instability does indeed exist, not only as it probably did in the October 87 storm but also in other damaging windstorms that have been investigated since. Other researchers found that flows resembling sting jets could be created by air descending into a region of stronger pressure gradient as it travels around the cyclone centre. The resulting dynamical forces occur on a larger scale – and hence are better known. These researchers suggest that sting jets might thus be generated without needing either evaporative cooling or the smaller-scale dynamical instability.
Weather forecasts have improved a great deal over the past three decades and if the October 87 storm were to be repeated, there would be useful warnings a few days in advance. The real challenge now is to predict whether there will be a small core of extreme winds due to a sting jet and just where and when it will occur. It is doubtful that the large-scale forces alone can be adequate to reliably predict the intensity and location of the most damaging winds associated with a sting jet.
An emerging view is that the greatest damage occurs in a sting jet due to a combination of all of the factors above: the large-scale forces set the scene and the smaller-scale processes account for the focused nature of the area of greatest damage. This means that weather forecast models need to have very high resolution in order to properly resolve the small-scale processes. Also, if cooling by evaporation is confirmed to be important, there will need to be improvements in the way in which clouds are represented in these models. The representation of cloud processes is a key challenge not only in weather forecast models but in climate models too.